© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 3151
    Prev
    Next
    INDEX
    Explanations

    Romanized versions of country names

    oai_token-act-pair · gpt-3.5-turbo
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    373
    +0.15
    0.7%
    314
    +0.11
    0.5%
    232
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.15
    0.03
    232
    +0.11
    0.03
    373
    +0.09
    -0.01
    Negative Logits
     ransom
    -0.66
     Shades
    -0.64
    ANK
    -0.62
    à¨
    -0.61
     Deliver
    -0.60
     Conquer
    -0.59
    è¦ļéĨĴ
    -0.57
    GROUND
    -0.57
     Courage
    -0.57
     à¨
    -0.56
    POSITIVE LOGITS
    odus
    1.07
    posure
    1.01
    cellent
    0.97
    actly
    0.97
    ample
    0.96
    avier
    0.95
    ercise
    0.93
    peria
    0.93
    cellence
    0.91
    otic
    0.91
    Activations Density 0.037%

    No Known Activations